Melting pot or salad bowl: Some evidence from U.S. investments abroad

We estimate the number of foreign-origin persons in the United States classified by their country of origin from census data in 1970, 1980, 1990 and 2000. We find, both in cross-sectional tests and in panel data tests, that the size of the foreign-origin group from a country living in the U.S. is positively correlated with U.S. investments in that country. This national origin bias is strong for direct (FDI) and modest for indirect (equity holdings) investments. The results continue to hold even after controlling for the "fundamentals" hypothesized to affect foreign investments. The other economic geography variables of a country-physical distance from the U.S., race, language and religion-do not seem to affect US investments in that country.